INTERNATIONAL RESEARCH JOURNAL OF MODERNIZATION IN ENGINEERING TECHNOLOGY AND SCIENCE

Abstract

ABSTRACTCybersecurity is an issue of most significance today, and it is important to detect and mitigate such phishing Uniform Resource Locators (URLs). One approach for detecting phishing URLs using machine learning techniques is presented in this research. The proposed system uses a combination of benign and phishing URLs as a dataset to train various machine learning models. Feature engineering can be used to extract important attributes from URLs, while different algorithms are combined in order to find the best model. To assess how efficacious the suggested solution is, this study measures its performance by means of accuracy, precision, recall, and F1 score. As illustrated by the findings, the authors have successfully applied the machine-learning-based method for distinguishing between safe and fake URLs, which can be utilized as valuable tools leading to better online security measures. This article offers an intelligent algorithm that detects phishing URLs, thereby contributing to strengthening the cybersecurity framework in use presently.INTRODUCTIONIn the contemporary landscape of evolving online threats and the incessant surge in cyberattacks, safeguarding digital environments has become an imperative task. A significant facet of this defense strategy involves the early identification and mitigation of phishing Uniform Resource Locators (URLs), which serve as gateways for cyber threats. Phishing URLs often disguise themselves within the vast expanse of the internet, necessitating advanced and adaptive detection mechanisms.This research introduces a comprehensive approach to address the challenges posed by phishing URLs through the application of machine learning techniques. Leveraging the power of data-driven models, we present a system that utilizes a carefully curated dataset containing both benign and phishing URLs. The primary objective is to train and evaluate various machine learning models to effectively discern between harmless and phishing web addresses.Our methodology encompasses feature engineering, a critical aspect in extracting relevant characteristics from URLs, which serves as the foundation for model training. This study conducts an in-depth evaluation of several machine learning methods to figure out the best effective approach to stable and precise recognition. The performance evaluation of the system is carried out using key metrics such as accuracy, precision, recall, and F1 score, providing nuanced insights into the system's efficacy.Beyond the quantitative assessment, this study delves into the qualitative aspects of the proposed solution, emphasizing its robustness in handling the dynamic nature of phishing URL patterns. The results obtained not only showcase the effectiveness of the machine learning-based approach in distinguishing between benign and phishing URLs but also highlight its potential as a valuable and intelligent tool for enhancing online security measures.By contributing a novel and adaptive solution for the proactive identification of phishing URLs, this research aligns with the ongoing efforts to fortify cybersecurity infrastructure. As cyber threats continue to evolve, our intelligent system demonstrates a forward-looking approach, empowering organizations and individuals with an advanced defense mechanism against the ever-growing sophistication of online threats.

Paper File to download :